nature-ref-verifier

对学术文献逐条执行多源交叉验证，逐字段对比作者、标题、年份、卷期、页码，标记卷年/DOI年冲突、作者顺序异常、页码偏差等问题，输出结构化验证报告。可批量处理整篇论文/开题报告的参考文献列表，也可单条校验，支持与 Zotero 同步修正。

Quality

72%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Low

Low-risk findings worth noting

Fix and improve this skill with Tessl

tessl review fix ./skills/nature-ref-verifier/SKILL.md

Quality

Content

77%

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

A well-structured, highly actionable instruction skill with a clear multi-step workflow and explicit validation checkpoints; its main weaknesses are a redundant ASCII diagram that inflates token use and an existing reference file that is not linked from the body.

Suggestions

Remove or shrink the ASCII strategy diagram, since the Step 2 prose already conveys the same source/compare/rate flow.

Link to references/common-patterns.md from the Step 3 comparison section and move the duplicated case examples there so the body stays a concise overview.

Consider pulling the large Step 3 case-list tables into the reference file, leaving only the severity definitions and one example per tier inline.

Dimension	Reasoning	Score
Conciseness	Mostly efficient and free of generic concept explanations, but the 33-line ASCII strategy diagram restates the Step 2 workflow text and the case tables pack in many concrete examples, so it could be tightened; not fully lean per the score-3 anchor.	2 / 3
Actionability	Gives concrete, copy-paste-ready guidance: the Crossref endpoint 'https://api.crossref.org/works/<DOI>', 404/200 decision rules, a >20 条 batch threshold with 10–15 per group, and specific Zotero write paths (localhost:23119 read-only, pyzotero web mode). For an instruction-only skill the guidance is fully actionable.	3 / 3
Workflow Clarity	Clear 5-step sequence with explicit validation checkpoints — the cheapest-first Step 2.0 DOI check with 404/200 branching, and the Step 4 confidence tiers — plus batch feedback guidance (verify item is in library before fixing, 429 backoff), matching the score-3 anchor.	3 / 3
Progressive Disclosure	Sections are well-organized and a real references/common-patterns.md bundle exists one level deep, but the body never links to it and overlaps with its content in Step 3's case tables, so the reference is not clearly signaled and the body is partly monolithic.	2 / 3
	Total	10 / 12 Passed

Description

67%

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A specific, distinctive description written in correct third-person voice, but it omits explicit 'Use when...' trigger guidance, leaving the 'when to invoke' question implied rather than stated. The description also lacks natural user-facing trigger terms, which constrain trigger_term_quality and completeness.

Suggestions

Add an explicit 'Use when...' clause to the description, e.g. 'Use when verifying a bibliography before paper/proposal submission, after reviewer comments about citations, or during Zotero library health checks.'

Surface a natural trigger term inside the description (e.g. '校验文献' or 'verify references') so the description itself carries user-facing keywords rather than only in the body.

Keep the concrete action list but tighten it to one line so the trigger guidance fits within the description without padding.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete actions such as '逐条执行多源交叉验证', '逐字段对比作者、标题、年份、卷期、页码', '标记卷年/DOI年冲突' and '输出结构化验证报告', matching the score-3 anchor of multiple specific concrete actions.	3 / 3
Completeness	It clearly answers 'what' (multi-source cross-validation, field comparison, structured report) but has no explicit 'Use when...' trigger guidance, so per the judging guidelines completeness is capped at 2.	2 / 3
Trigger Term Quality	The description names the domain (学术文献/参考文献) and operation nouns like '单条校验' and '同步修正', but lacks natural trigger phrases a user would actually say; the natural triggers ('verify references', '校验文献') live only in the body, not the description.	2 / 3
Distinctiveness Conflict Risk	The academic-reference cross-validation niche with DOI/Zotero specifics is clearly distinct and unlikely to trigger the wrong skill, matching the score-3 anchor.	3 / 3
	Total	10 / 12 Passed

Validation

100%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository: Yuan1z0825/nature-skills
Path: skills/nature-ref-verifier/SKILL.md
Commit: 703c150

Reviewed: 4 days ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.